[PATCH AUTOSEL 6.6 41/59] ALSA: hda: Disable jack polling at shutdown

[Sasha Levin <sashal@kernel.org>]

From: Takashi Iwai <tiwai@suse.de>

[ Upstream commit 1adcbdf54f76e1004bdf71df4eb1888c26e7ad06 ]

Although the jack polling is canceled at shutdown in
snd_hda_codec_shutdown(), it might be still re-triggered when the work
is being processed at cancel_delayed_work_sync() call.  This may
result in the unexpected hardware access that should have been already
disabled.

For assuring to stop the jack polling, clear codec->jackpoll_interval
at shutdown.

Reported-by: Joakim Zhang <joakim.zhang@cixtech.com>
Closes: https://lore.kernel.org/20250619020844.2974160-4-joakim.zhang@cixtech.com
Tested-by: Joakim Zhang <joakim.zhang@cixtech.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20250623131437.10670-2-tiwai@suse.de
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

LLM Generated explanations, may be completely bogus:

**Backport Status: YES**

This commit should be backported to stable kernel trees. Here's my
extensive analysis:

## Bug Fix Analysis

The commit fixes a **race condition** during codec shutdown where jack
polling work could be re-triggered even after attempting to cancel it.
This is evident from the commit message: "it might be still re-triggered
when the work is being processed at cancel_delayed_work_sync() call."

The race condition occurs because:
1. `cancel_delayed_work_sync()` waits for the current work to complete
2. While waiting, if the work function (`hda_jackpoll_work`) is
   executing, it checks `codec->jackpoll_interval` at line 635 of the
   original code
3. If `jackpoll_interval` is still non-zero, the work reschedules itself
   with `schedule_delayed_work(&codec->jackpoll_work,
   codec->jackpoll_interval)` at line 641
4. This creates a situation where hardware access occurs after shutdown
   has supposedly completed

## Code Change Analysis

The fix is **minimal and surgical** - just one line added:
```c
codec->jackpoll_interval = 0; /* don't poll any longer */
```

This is placed **before** the `cancel_delayed_work_sync()` call,
ensuring that even if the work function is currently running, it won't
reschedule itself because the interval check at the beginning of
`hda_jackpoll_work()` will fail.

## Impact Assessment

1. **User-visible bug**: Yes - unexpected hardware access during
   shutdown can cause system hangs, crashes, or hardware-related issues
2. **Fix size**: Extremely small - single line addition
3. **Risk assessment**: Very low - setting a value to 0 before canceling
   work is a safe operation
4. **Subsystem criticality**: ALSA HD-Audio is widely used, but this
   specific fix is in the shutdown path
5. **Testing**: The commit has been tested by the reporter (Joakim
   Zhang)

## Stable Tree Criteria

This commit meets stable tree requirements:
- Fixes a real bug (race condition causing unexpected hardware access)
- Minimal change (1 line)
- Clear fix with obvious correctness
- No new features or architectural changes
- Only affects the shutdown path, limiting scope
- Has been tested by the bug reporter

The fix prevents potential system instability during shutdown, making it
an important candidate for stable backporting.

 sound/pci/hda/hda_codec.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/sound/pci/hda/hda_codec.c b/sound/pci/hda/hda_codec.c
index 2d4b7527b840..80c3084189b0 100644
--- a/sound/pci/hda/hda_codec.c
+++ b/sound/pci/hda/hda_codec.c
@@ -3038,6 +3038,7 @@ void snd_hda_codec_shutdown(struct hda_codec *codec)
 	if (!codec->core.registered)
 		return;
 
+	codec->jackpoll_interval = 0; /* don't poll any longer */
 	cancel_delayed_work_sync(&codec->jackpoll_work);
 	list_for_each_entry(cpcm, &codec->pcm_list_head, list)
 		snd_pcm_suspend_all(cpcm->pcm);
-- 
2.39.5